Stable Diffusion Prompt Photo

AI prompts are the text instructions that guide AI image generation models like Stable Diffusion on what type of image to generate. Crafting effective prompts is key to producing high-quality AI art.

This article will provide prompt examples and tips to help you get started with Stable Diffusion.

What Makes a Good Prompt?

A good Stable Diffusion prompt should:

  • Clearly describe the desired image – Be specific about the content, style, composition, etc.
  • Provide relevant context and details – Give the AI enough information to work with.
  • Use formatting for clarity – Organize prompts with commas, line breaks, etc.
  • Limit prompts to 75-100 tokens – Shorter prompts tend to work better.

Prompt Structure

Stable Diffusion prompts generally follow this structure:

[Content], [Style], [Composition], [Quality and Rendering]

Here is an explanation of each component:

Content

The content describes what you want to see in the image. This should be 1-2 sentences summarizing the main subject and scene.

For example:

A painting of a girl playing near a pond in a forest

Style

The style section indicates the desired artistic style. Reference specific artists, genres, art movements, etc.

For example:

in the style of impressionist oil paintings by Claude Monet

Composition

The composition component provides details on framing, orientation, depth of field, etc.

For example:

soft lighting, depth of field, rule of thirds composition

Quality and Rendering

This section describes the desired quality and rendering of the final image. You can specify things like:

  • Level of realism vs abstraction
  • Brightness/contrast
  • Color palette
  • Medium (paint, pencil drawing, etc.)

For example:

intricate realism, sharp focus, vibrant spring colors, digital art

Prompt Example Breakdowns

Below are some full prompt examples for Stable Diffusion broken down by section.

Landscape Photography Prompt

A misty forest in autumn, red and yellow leaves on the trees, evergreen forest, mystical morning light shining through the trees, peaceful mood

cinematic lighting, depth of field, vibrant warm colors, intricate details, by Wes Anderson and Hayao Miyazaki  

wide landscape orientation, rule of thirds composition, epic perspective

intricate realism, ultra-detailed, sharp focus, high resolution, digital painting

This prompt generates a highly detailed, vivid landscape photo with specific stylistic influences.

Character Portrait Prompt

A wise fantasy wizard with a long white beard, purple robes, and mystical staff, casting a spell over a magical glowing orb 

intricate ink drawing style, flowing lines, highly textured

close-up portrait, rule of thirds composition, dramatic side lighting

intricate realism, ultra-detailed, sharp focus and textures, high resolution

This prompt generates a textured, dramatic portrait of a wizard character in ink drawing style.

Advanced Prompting Techniques

Here are some advanced techniques you can use to further refine your Stable Diffusion prompts:

Negative Prompts

Add negative prompts with - to specify what you don’t want in the image. For example:

-ugly, -poorly drawn, -bad anatomy

Sampling Methods

Indicate a specific sampling method like k_euler a or k_dpmpp_2m to control the diffusion process.

Aspect Ratio

Set a specific aspect ratio like ar 16:9 for panoramic images.

Image Count

You can generate multiple variations by setting n=4 to create 4 images.

Prompt Inspiration

Stuck on ideas? Browse AI art communities like PromptHero and Lexica to discover new prompts. You can also generate prompts automatically using AI assistants.

The key is to keep experimenting until you find prompts that produce images matching your creative vision!